Exploratory Analysis
Project title
Gender-based Wage Gap in Professional Sports
Charlie Bond, Sirinya Hiruvinya, Aleah Rosner
Date
Spring 2024
Abstract
Our main goal with this project is to explore the difference in salary between NBA players and WNBA players, which is important because it points to a larger issue of gender inequality that spans nearly every career path. To address this question, we will be comparing the salaries of professional players across both leagues and using our dataset to analyze trends in salary vs games played, points scored, and other factors. When doing this, it is important to consider that this data spans a short time frame (2016-2017), which needs to be taken into account when making conclusions from our analysis; however, our goal of bringing light to the issue of the wage gap by applying it to the popular world of professional sports still stands.
Keywords
NBA, WNBA, wage gap, gender inequality, salary
Introduction
A few research questions we hope to explore using this data set include: How does statistical basketball performance, such as minutes played, points scored, rebounds, assists, steals, and blocks, differ between the WNBA and NBA? How do salary-to-performance ratios (salary per point, salary per rebound, salary per assist, etc.) compare between male and female basketball players? To what extent does salary vary between WNBA and NBA players when compared to their performance? Through this analysis, our intention is to examine how a broad issue of gender inequality in the workplace is applied to a niche such as elite sports. This is important because many people don’t acknowledge the fact that although playing at the most elite level of basketball is a rare feat for both men and women, the difference in pay between the men’s and women’s professional leagues is palpable. More broadly, we aim to bring awareness to the wage gap as a whole by introducing it in relation to a very popular sport. Acknowledging the wage gap is incredibly important because only after people are aware of it can changes begin to be made. By relating the wage gap to sports, it is possible that this issue can reach a wider audience and more people might be inspired to advocate for salary transparency in their workspace and careers.
The Dataset
\(Where was the data found?\)
The data was found by googling “WNBA salary dataset” and “NBA salary dataset”. The data we found for the Women’s league is from herhoopstats.com, and the NBA data was found on gigasheet.com. we also found data from 2016 by googling “wnba nba salary gap csv”.
\(Who Collected the data?\)
The WNBA data was collected by technologists at Sportradar, a sports technology company affiliated with HerHoopStats. The NBA data was uploaded by a community contributor who put the data in CSV format, but the information is available to the public by the NBA. The 2016 data was found on github.
\(How was the data collected?\)
The data is made open to public, and was collected and formatted into a csv file by technologists from companies such as sportradar.
\(Why was the data collected?\)
Typically, this data is collected for use by the media (such as sports announcers) and for sports betting purposes.
\(How many observations (rows) are in your data?\)
The WNBA has information on 130 players, and the NBA dataset has information on over 500 players. Combined, there are 630 observations.
\(How many features (columns) are in the data?\)
The original datasets had 27(WNBA) and 52(NBA) features, but we will condensed these to the 10 features we felt were most relevant to our analysis.
What, if any, ethical questions or questions of power do you need to consider when working with this data?
Because the salary data for WNBA and NBA players is available for the public to access, there aren’t as many ethical questions to be addressed in terms of anonymity. However, in order to avoid potential bias in our analysis and in other people’s exploration of the data, it might be a good idea to refrain from using specific players’ names in our analysis. This is especially important as many people already have preconceived notions about the salaries of the NBA compared to the WNBA, so it is important to ensure that our analysis and presentation of the data reflects only the numbers, and is not influenced by previous assumptions about the relative popularity of the two leagues.
What are possible limitations or problems with this data?
With our ultimate goal being to demonstrate the clear difference in salaries between men and women, one major drawback of our data is that it only showcases this disparity in one part of one very niche industry. Additionally, comparing the data we have on the wage gap in other industries (such as from the Forbes analysts mentioned above) to the data we’ve found in professional basketball leagues, the sports figures we will be analyzing demonstrate an extreme wage gap that is significantly larger than almost any other industry. For this reason, it’s important that we acknowledge that the differences we see between the WNBA and the NBA are more stark than in other areas of work. Additionally, the data we have on the WNBA has about half the entries (rows) compared to the NBA. Although we have adequate numbers of data points for both leagues, this difference could also lead to a slightly more accurate portrayal of one league compared to the other simply because more data is available. Keeping in mind both the extremity of the wage gap in this industry specifically and the difference in data points is important so that we don’t accidentally end up spreading misinformation, which could lead to a reversal of the progress that has been slowly but steadily made to close this gap.
Implications
By understanding the disparity evident through these statistics, data analysts and technologists could develop AI or data analysis tools that could assess player value and their contribution to the games. This would make it easier to objectively measure a player’s real talent and provide female players with more fair pay. This helps reduce biases and provides evident numerical data that could be used to determine a player’s salary.
Designers could take part in creating more compelling campaigns and narratives to highlight the strengths, achievements, and stories of female players in the WNBA. Designers play a critical role in shaping how sports are marketed to the general public, and through strategic and inclusive branding, they could challenge stereotypes and empower women in the sports industry. This will garner more attention for women in sports and WNBA games, which could drive more revenue and empowerment for the industry.
Our analysis could also highlight the severity of gender inequality to policymakers; policymakers could implement policies such as the pay transparency initiative for employers, invest more in women’s sports and representation, and monitor the pay gap in different industries. If technologists, designers, and policymakers work together, they could reduce the gender wage gap by coming up with systems and policies that objectively measure an individual’s talents and contributions that determine their wage.
Limitations & Challenges
While the datasets are ideal in size and readability, they were collected from different sources. as we discussed in class, biases and inaccuracies can skew or alter the accuracy of data, and coming from two different sources means that they might be affected by different biases. Additionally, given how niche the professional sports industry is, it is possible that the trend in the wage gap in this area of work differs from that of other industries. This poses an additional challenge for applying our analysis to current trends in the wage gap.
More broadly, one challenge with using this data is that it only addresses one niche of the gender pay gap, which may be difficult to generalize to an entire population. These datasets only include the salaries and pay data of professional basketball players from a very short time period (2022-2024). In order to address this, we can acknowledge this specificity in our discussion of our data analysis, or we could implement other sources (such as news articles) that point out this pay disparity in more general terms.
Summary Information
## [1] 8416599
## [1] 130795.7
## [1] 8285803
## [1] 48070014
## [1] 241984
## [1] 47828030
## [1] 5849
## [1] 64154
## [1] 58305
## [1] 11637.74
## [1] 6447.836
## [1] 5189.904
## [1] 850078.4
## [1] 18532.17
## [1] 831546.3After performing a preliminary analysis of the data and calculating a few values, the difference between female and male players’ earnings is astonishing. Beginning with the average salary within each league, we found that the average salary in the NBA was over $8.4165987^{6}, while in the WNBA, the average salary was a mere $1.3079572^{5}, which is a difference of $8.285803^{6}. Taking a closer look, we first compared the highest and lowest salaries between the leagues. We found that the maximum NBA salary (48,070,014) was $4.782803^{7} greater than the highest WNBA salary. Interestingly, the lowest NBA salary (5,849) was $ 5.8305^{4} less than the lowest WNBA salary of 64,154. We also compared salaries relative to performance metrics such as salary per point and salary per minute played. Although on average, WNBA players did tend to score less in games, the money they earned per point and minute played was still significantly less than their male counterparts. On average, NBA players made 850,078.40 per point, whereas WNBA players only earned 18,532.70 for every point they scored (a difference of 831,546.30). Additionally, NBA players earned 11,637.74 dollars for every minute of play, while female players in the WNBA only made 6,447.84 dollars (a difference of $5,189.90).
Table
This was included to give some general impressions on differences between each league. The stats include Average Playtime per game, and Average Salary per Minute, which gives an idea of how the players are compensated when accounting for differences in the structure of each league. The Proportion of Games Played out of Games Started gives an idea of how reliably a player is playing a full game out of all their court appearances, which makes for important context when considering the pay differences between each league.
Chart 1
##Author: Charlie Bond
This chart takes looks at average salaries for each league and divides them by various performance statistics. The goal of this is to compare the salaries between the leagues accounting for performance and what is rewarded in each league. The ultimate purpose is to show the sheer scale of difference in how players get rewarded in each league, accounting for differences between the leagues. Some noteworthy observations paint the stark difference between each league: the average salary per point for the WNBA cannot be observed on the scale of this bar plot, but is barely visible for each of the other performance metrics. As such, this graph serves a purpose of underscoring the sheer scale in difference between the two leagues. =======
Chart 2
Chart 3
Author: Aleah Rosner
This chart displays the change in average salary in each league between 2016 and 2024. The purpose for showing this is to examine and compare the trend in calaries among the mens and womens leagues, which can help us begin to determine whether progress is being made to close the wage gap. In 2016, the average salary for both is 100% (or 1x) the amount that it was in 2016, which is why both graphs start at one. The fold change (amount of increase) was calculated using the formula (2024 avg salary)/(2016 avg salary). Using this, we determined that the WNBA’s average salary increased to around 1.25 times (125%) the salary from 2016, while the NBA’s average salary to 0.94 (or 94%) of its 2016 salary. This data is promising, as it shows that the WNBA’s salaries are, in general, on the rise.